This study will address two important and unresolved issues regarding the maintenance therapy of patients with bipolar affective illness. One concerns the minimum serum lithium concentration effective for the prevention or attenuation of future affective recurrences. We plan to compare and contrast two different levels of serum lithium (0.4 to 0.6 and 0.8 to l.0 mEq/liter) regarding their ability to protect against relapse, the presence of less severe affective symptoms, and their association with unwanted and toxic effects--such as impairment of renal and thyroid function and uncomfortable "side effects" (e.g., tremor, weight gain, and cognitive impairment). The second component of this proposal focuses on a group at particularly high risk for relapse, contrasting two maintenance regimens for patients who experience manic relapses despite lithium maintenance; continued lithium therapy alone versus lithium plus fluphenazine decanoate. Both studies will be double-blind and will involve a collaboration between the Massachusetts General Hospital and Long Island Jewish-Hillside Medical Center. The first will involve l75 subjects at the two sites over five years, the second 50 subjects at the two sites over five years.